The new-authority washout

Brokers treat a brand-new MC number as riskier than an established one. They are right that it is riskier. They are usually wrong about why.

The instinct is that a new carrier is a safety unknown, an untested operator more likely to put a truck in a ditch. When we tested that against the federal record, it did not hold. The real early-life risk is a different thing entirely, and once you name it correctly it is a sharper signal than the one most people are watching for.

One in seven new authorities washes out

Every carrier that registers for interstate operating authority enters an 18-month probation called the New Entrant Safety Assurance Program: it must pass a safety audit to keep the authority. A large and remarkably stable share never make it through.

A fresh MC number is, in effect, a coin with a one-in-seven chance of not surviving its first 18 months as a registered carrier.

Almost none of it is a safety shutdown

Here is the part that reframes the whole thing. When we split those new-entrant revocations by reason, the overwhelming majority are not carriers caught operating dangerously. They are carriers that never engaged with the audit at all.

This is a specific, narrow kind of revocation, and it is worth being precise about which one. It is not an insurance lapse, and it is not a for-cause safety shutdown of an operating carrier. It is the new-entrant audit washout: a fresh registrant that never completed, or never responded to, the mandatory audit. Many of these authorities were speculative or never turned into a real operating business in the first place.

The crash part of the story is a myth

If new carriers washed out because they were dangerous, you would expect them to crash more. They do not. The raw record makes older carriers look far more crash-prone, but that is almost entirely because older carriers run more trucks. Put the two on the same footing, crashes per truck rather than per carrier, and the picture inverts: a carrier's first 18 months are among the lowest-crash stretches of its life, not the highest.

We pressure-tested that the way it is most likely to be wrong. Crash reports lag, and the lag falls hardest on the newest carriers, so an incomplete-data artifact could have been hiding a higher young rate. Recomputing on a window old enough to be fully reported did not close the gap, it widened it slightly. Per truck, new authority is genuinely the safer end of the curve.

So the early-life risk is real, but it is a business-survival and legitimacy risk, not a crash risk. The question a fresh MC number actually raises is not "will this carrier be unsafe," it is "is this a real, still-operating business that will complete its federal audit, or a registration that never became a company."

Where the washed-out carriers go

Revoked does not always mean gone. A share of carriers that lose a new authority simply re-register under a fresh USDOT number and start the clock over, which is the reincarnation pattern we cover in a separate report. So the washout and the chameleon story are, in part, the same carriers seen twice: once when the first authority fails its audit, and again when a new one appears in its place.

How we did it without fooling ourselves

1. We controlled for fleet size. Counting whether a carrier "ever had a crash" makes big, old fleets look dangerous purely because they have more trucks and more years. We measured crashes per power unit instead, which is what actually compares a new carrier to an old one.
2. We controlled for time at risk. A ten-year carrier had ten years to accumulate a crash; a six-month carrier had six months. We used a fixed recent window and credited each carrier only for the time it actually existed inside it.
3. We separated the kinds of revocation. "Revoked" means several different things in this data. We isolated the new-entrant audit washout from insurance-lapse revocations and from for-cause safety shutdowns, and only the audit washout concentrates in the first 18 months. For-cause safety failures, by contrast, accumulate across a carrier's whole life.
4. We checked the biases run the safe way. Failed carriers stay in the record, and crash reporting lags most on the newest carriers, both of which would tend to make young carriers look worse, not better. The finding that they are not more crash-prone holds up against the pressure, not because of it.